The following repo is an attempt to implement English-Italian machine translation. The transformer architecture based on the paper "Attention is All you need" implemented from scratch using Pytorch/Python and no other high level API(nn.Transformer etc.). We utilised Autokenizer from huggingface to perform the task of tokenization. Following dataset was utilised for training https://huggingface.co/datasets/g8a9/europarl_en-it.
( for training loop/complete notebook kindly please open issue/PR) cheers!
